In this project, I analyze my tracks ‘desmond-l-1’ and ‘desmond-l-2’ compared to the Computational Musicology 2025 corpus using Essentia features such as danceability, energy, valence, and instrumentalness.
I curated my tracks using generative AI tool Stable Audio. The prompts I used were based on two of my favorite artists: KI/KI and Marlon Hoffstadt. Both DJs are known for their high energetic music, combining acid and trance with techno.
The table below presents the extracted Essentia feature values for my generated tracks.| filename | approachability | arousal | danceability | engagingness | instrumentalness | tempo | valence | ai |
|---|---|---|---|---|---|---|---|---|
| desmond-l-1 | 0.2629817 | 4.478108 | 0.2859525 | 0.4156072 | 0.6434987 | 135 | 3.936315 | TRUE |
| desmond-l-2 | 0.2929443 | 5.076702 | 0.3010519 | 0.5524329 | 0.4989389 | 73 | 4.316221 | TRUE |
This graph explores the relationship between arousal and valence, with the danceability represented by the color.
My tracks, desmond-l-1, and desmond-l-2 score high in arousal and valence, meaning they are energetic and also emotionally positive. This aligns perfect with the sound of KI/KI and Marlon Hoffstadt.
Compared to the class corpus, my tracks exhibit a higher combination of energy, emotional positivity, and danceability. This suggests that they are:
Chroma-based Self-Similarity: In this Self-Similarity Matrix, we identify gridlike patterns in the track, suggesting repeating harmonic structures, especially in the area from 50s-100s, as well as near the end. Overall the track we see repeated patterns, suggesting recurring motifs.
Timbre (MFCC) Self-Similarity: Here we see bright edges, implying that the intro and outro are similar. In the middle, we get segments where we have darker regions, especially from 90s-140s, meaning variations in timbre. We see patterns that reoccur meaning that we have repeating timbral elements throughout the track.
Chromagram: In the Chromagram, we observe that the A note has a prominent frequency, especially from 55s and on. We also observe that between 00s-55s and 110-150s we have a structure with a presence of different pitches, meaning we have a varied harmonic structure.
Cepstrogram:: We observe a quite stable cepstrogram throughout the track, meaning that there are no drastical changes throughout the track.
Track 1:
Keygram: In this graph we see obvious parts where the track is split into key areas. From 55s-100s we see brighter key areas, suggesting stronger and more persistent keys in that time period. This pattern is also observed from 150s-180s. By listening to this track, we hear the transitions that are observed in the keygram.
Chordogram: Similarly to the keygram, we observe the same structures in the same time periods. However, from 0s-55s we observe a different pattern, where the sections are more all over the place. This suggests frequent chord changes and modulation between keys. This structure is also observed from 100s-150s. The areas 55s-100s and 150s-180s have a more stable pattern, suggesting recurring chord progressions.
Track 2:
Keygram: Here we observe a more distributed key activation compared to the first track. The lack of bright areas also suggest more key shifts, opposed to the patterns observed in the first track. This means that the second track is more varied harmonically. No key dominates over a longer time period.
Chordogram: Compared to the keygram, the chordogram has more persistent bright regions, meaning that certain chord progressions last for longer periods. Overall, this track is more complex and chaotic than the first track. While listening to the track, I can confirm this. This track has way more frequent modulations and less structure, feeling all over the place.
Tempogram Track 1: TBC
Tempogram Track 2: TBC
Track 1:
Energy Novelty: TBC
Spectral Novelty: TBC
Track 2:
Energy Novelty: TBC
Spectral Novelty: TBC